Effects of Five State Prekindergarten Programs on Early Learning

By W. Steven Barnett, Kwanghee Jung, Vivian Wong, Tom Cook, and Cynthia Lamy
(2007)

This NIEER study of high-quality prekindergarten programs in five states reveals significant improvement in children's early language, literacy and mathematical development, producing somewhat larger average estimated effects than the national randomized trial of the Head Start program. However, there are numerous factors that may have led to underestimation of Head Start's effects. As the state pre-K programs in this study are of relatively high quality, estimates cannot be generalized to all state pre-K programs.

The study finds that children attending state-funded pre-K programs in the five states (Michigan, New Jersey, Oklahoma, South Carolina, and West Virginia) gained significantly regardless of ethnic background or economic circumstances. Substantial variation in estimates among the states, but this variation can be due to interstate differences in access to early education outside state pre-K as well as to differences among state pre-K programs. Key findings were:

• Applying one common model to data from all states, estimated effects were statistically significant for New Jersey and Oklahoma, which both had effect sizes of about .33. Averaged across all states the effect size was .18.
• The estimated effect of state-funded prekindergarten on children's scores on the Woodcock-Johnson-III Applied Problems subtest was statistically significant for all four states in the pooled analyses. The average effect size across all four states was .43.
• The estimated effects of the five state-funded pre-K programs on children's Print Awareness were statistically significant for each state program in the pooled analyses. The estimated effect sizes ranged from .46 to 1.10 with an average of .74.
• When pre-K effects were estimated individually for each state, estimates were modestly smaller on average and more variable.

View Full Report